What statistics should we use in evaluating the effectivity time? Our high school experimental study is about insecticide. Specifically, we need to know the effectiveness of guyabano (Annona muricata) seeds as dog tick reducer solution and which is the best treatment. But we do not know what statistics to be used in knowing the effectivity time. What should we use? 
 A: There are several very important considerations in the design stage: first, what is the sampling frame or the unit analysis for your study? Next, what is the outcome you intend to measure? And third, what are the treatment conditions your class thinks should be measured? With respect to the first consideration can we assume that your sampling frame is dogs with ticks? Next, your class could count the number of ticks for specific areas on a dog's skin before (or "pre") application of the solution as well as after (or "post") application. Note that it would be useful to count these ticks for areas roughly the same size and location on each dog. With respect to treatment conditions, you mention "effectivity time." Is this referring to the time the solution is left on the dog's coat, e.g., 5 minutes, 10 minutes, and so on? Next, is it possible to adjust the strength of the solution, e.g., 5%, 10%, etc. Given that, pre-post tick counts would be the outcome, time on coat and strength of solution would be the treatment conditions. Making decisions like these in advance is a very good idea.
With a design such as outlined above, a straightforward statistical test would be to develop an Analysis of Variance or ANOVA where the response you are testing is the pre-post average counts of ticks as a function of the time the solution was left on the coat and the strength of the solution. 
